CONDITION MONITORING AND PROTECTION OF TRANSMISSION LINE USING PMU

Abstract

Naive Bayes (NB) Classifier has been implemented for fault detection and type of fault classification in transmission line. The Naïve Bayes classifier is based on Bayes conditional probability theorem. Synchronized fault current is taken from two PMU placed at two alternate buses. The magnitude and angle of all the three phase current are considered for fault classification of transmission line. First the line on which fault has occurred has been identified using Naive Bayes classifier. Then type of fault is classified using maximum value of phase current. Since, PMU gives RMS value as output and the faulty phase shows maximum value of RMS current. So the maximum value of each phase current is used in Naïve Bayes classifier for fault classification. Ten different types of fault have been classified. It classifies the fault very accurately. The classifier is very simple to understand and use. It takes very less time for training and testing.